Browse

Automatic prediction of computational resource consumption for efficient task migration in cloud

DC Field Value Language
dc.contributor.advisor백윤흥-
dc.contributor.author권용인-
dc.date.accessioned2017-07-13T07:07:01Z-
dc.date.available2017-07-13T07:07:01Z-
dc.date.issued2015-02-
dc.identifier.other000000024860-
dc.identifier.urihttps://hdl.handle.net/10371/119051-
dc.description학위논문 (박사)-- 서울대학교 대학원 : 전기·컴퓨터공학부, 2015. 2. 백윤흥.-
dc.description.abstractIn order to accommodate the high demand for performance in smartphones, mobile cloud computing techniques, which aim to enhance a smartphone's performance through utilizing powerful cloud servers, were suggested. Among such techniques, execution offloading, which migrates a thread between a mobile device and a server, is often employed. In such execution offloading techniques, it is typical to dynamically decide what code part is to be offloaded through decision making algorithms. In order to achieve optimal offloading performance, however, the gain and cost of offloading must be predicted accurately for such algorithms. Previous works did not try hard to do this because it is usually expensive to make an accurate prediction.

Moreover, existing schemes completely ignore the costs of cloud resources by assuming that idle servers are always available for free of charge. These unrealistic assumptions make each server run only a small load to achieve the guaranteed high offload performance. Therefore, these schemes cannot be applied to real-world commercial clouds which aim to minimize the operation costs by maximizing the server throughput, and then charge users for their resource usage.

Thus in this dissertation, I present Mantis, a framework for predicting the Computational Resource Consumption(CRC) of Android applications on given inputs accurately, and efficiently. CRC synergistically combines techniques from program analysis and machine learning. It constructs concise CRC models by choosing from many program execution features only a handful that are most correlated with the program's CRC metric yet can be evaluated efficiently from the program's input. I apply program slicing to reduce evaluation time of a feature and automatically generate executable code snippets for efficiently evaluating features.
Using the techniques, I empirically show they enhance the performance of offloading. Lately, I propose CMcloud, a novel cost-effective mobile-to-cloud offloading platform, which works nicely under the real-world cloud environments. CMcloud minimizes both the server costs and the user service fee by offloading as many mobile applications to a single server as possible, while satisfying the target performance of all applications.
-
dc.description.tableofcontentsAbstract i
Chapter 1 Introduction 1
1.1 Mobile Execution Offloading . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Dynamic Code Partitioning . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Cost-effectivity of Mobile Execution Offloading . . . . . . . . . . . . 3
1.4 Dissertation Contributions and Outline . . . . . . . . . . . . . . . . . 4
Chapter 2 Mantis: Efficient Predictions of Execution Time, Energy Usage,
Memory Usage and Network Usage on Smart Mobile Devices
6
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2 Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.3 Feature Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.4 CRC Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.5 Predictor Code Generation . . . . . . . . . . . . . . . . . . . . . . . 15
2.5.1 Rationale . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.5.2 Slicer Challenges . . . . . . . . . . . . . . . . . . . . . . . . 17
2.5.3 Slicer Design . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.6 Implementations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.7 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.7.1 Evaluation Environment . . . . . . . . . . . . . . . . . . . . 24
2.7.2 Experiment Results . . . . . . . . . . . . . . . . . . . . . . . 26
2.8 Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.9 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Chapter 3 Precise Execution Offloading for Applications with Dynamic
Behavior in Mobile Cloud Computing 40
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.2 Background & Motivation . . . . . . . . . . . . . . . . . . . . . . . 41
3.2.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.2.2 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.3 f Mantis : Automatically generation of accurate and efficient performance
predictor for mobile execution offloading . . . . . . . . . . . . 48
3.3.1 Performance predictor generation overview . . . . . . . . . . 49
3.3.2 Profiler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.3.3 Predictor Generator . . . . . . . . . . . . . . . . . . . . . . 50
3.4 Dynamic code partitioning with predictor generated by f Mantis . . . 52
3.4.1 Architecture for our solver . . . . . . . . . . . . . . . . . . . 52
3.5 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.5.1 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.5.2 Evaluation Environment . . . . . . . . . . . . . . . . . . . . 55
3.5.3 Experimental results . . . . . . . . . . . . . . . . . . . . . . 56
3.6 Related work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Chapter 4 CMcloud: Cloud Platform for Cost-Effective Offloading of
Mobile Applications 73
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.2 Backgrounds and Limitations . . . . . . . . . . . . . . . . . . . . . . 75
4.2.1 Basic Offload Mechanisms . . . . . . . . . . . . . . . . . . . 76
4.2.2 Limitations of Existing schemes . . . . . . . . . . . . . . . . 77
4.3 CMcloud offloading . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.3.1 Design Goals . . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.3.2 Operation Model . . . . . . . . . . . . . . . . . . . . . . . . 80
4.3.3 Architecture Model . . . . . . . . . . . . . . . . . . . . . . . 82
4.4 CMcloud mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.4.1 Reference-model Server Profiling . . . . . . . . . . . . . . . 84
4.4.2 Performance Estimation . . . . . . . . . . . . . . . . . . . . 85
4.4.3 Performance Monitoring . . . . . . . . . . . . . . . . . . . . 92
4.4.4 Migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.4.5 Cost-aware Application Scheduling in Cloud . . . . . . . . . 94
4.5 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
4.5.1 Estimating Target CPI stack . . . . . . . . . . . . . . . . . . 96
4.5.2 Predicting Instruction Count . . . . . . . . . . . . . . . . . . 98
4.5.3 Cost Effectiveness with QoS requirements . . . . . . . . . . . 98
4.5.4 Offloading/migration Overhead . . . . . . . . . . . . . . . . 102
4.6 Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
4.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Chapter 5 Conculsion 105
초록 119
v
-
dc.formatapplication/pdf-
dc.format.extent3423027 bytes-
dc.format.mediumapplication/pdf-
dc.language.isoen-
dc.publisher서울대학교 대학원-
dc.subject모바일 컴퓨팅-
dc.subject성능 예측-
dc.subject.ddc621-
dc.titleAutomatic prediction of computational resource consumption for efficient task migration in cloud-
dc.typeThesis-
dc.contributor.AlternativeAuthorYongin Kwon-
dc.description.degreeDoctor-
dc.citation.pagesx, 119-
dc.contributor.affiliation공과대학 전기·컴퓨터공학부-
dc.date.awarded2015-02-
Appears in Collections:
College of Engineering/Engineering Practice School (공과대학/대학원)Dept. of Electrical and Computer Engineering (전기·정보공학부)Theses (Ph.D. / Sc.D._전기·정보공학부)
Files in This Item:
  • mendeley

Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.

Browse